- added
rhat
argument toglance()
andconverged()
to adjust convergence threshold, with default value of 1.05. Convergence no longer depends oness
. - removed
exclude_year
argument inbb_fit_survival_ml()
as failure to fit fixed year effect resolved with intercept adjustment. - provide warning to user if start of caribou year (
year_start
) not matching for survival and recruitment model fits inbb_predict_growth()
.
- Added
bb_predict_calf_cow_ratio()
bb_predict_recruitment()
now predicts adjusted recruitment, following methods in DeCesare (2012). Previous output ofbb_predict_recruitment()
can now be obtained by runningbb_predict_calf_cow_ratio()
.- Renamed
yearling_proportion_female
argument inbb_fit_recruitment()
tosex_ratio
. This matchesbbouretro
language and arguments inbb_predict_recruitment()
. - Changed default value of
include_uncertain_morts
argument inbb_fit_survival()
to TRUE. This matches default behaviour inbbouretro
. - Added article on 'Prior Selection and Influence'.
- Added example to "Analytical Methods" article.
- Moved internal data checking functions to be exported function in
bboudata
. - Updated bbtutorial vignette to mention the new data checking functions.
bb_fit_recruitment()
will error if there are any NA values in the PopulationName, Year, Month and Day columns.bb_fit_survival()
will error if there are any NA values in the PopulationName, Year, Month, StartTotal, MortalitiesCertain and MortalitiesUncertain columns.
Added 'Getting Started' vignette.
Added functions:
bb_predict_survival_trend()
bb_predict_recruitment_trend()
bb_plot_year_trend_survival()
bb_plot_year_trend_recruitment()
bb_predict_population_change()
bb_plot_year_population_change()
plot()
(plot traceplots for bboufit object)
Added arguments:
year_start = 4L
tobb_fit_survival()
andbb_fit_recruitment()
to set caribou year start month as integer - (default April)yearling_female_proportion = 0.5
tobb_fit_recruitment()
to set yearling sex ratioyear_trend = FALSE
tobb_fit_survival()
andbb_fit_recruitment()
to fit model with year trendinclude_random_effects = TRUE
tocoef()
/tidy()
to provide possibility of removing individual random effect estimates
- Switch MCMC engine from Jags to Nimble.
- Added recruitment/survival Maximum Likelihood fit functions:
bb_fit_survival_ml()
bb_fit_recruitment_ml
- Added summary methods for Maximum Likelihood output with class 'bboufit_ml'.
- Changed how priors are passed to fit functions (e.g.
priors = c(b0_mu = 4, b0_sd = 2)
instead ofpriors = list(b0 = "dnorm(4, 2^-2)")
). - Added argument
sig_fig
tocoef()
/tidy()
/predict()
methods to round estimates and lower/upper bounds to set significant figures.
bb_predict_growth()
now handles different years and doesn't return month.- Adjusted default priors.
- Soft-deprecated
bb_predict_lambda()
forbb_predict_growth()
. - Added
bb_plot_year_growth()
. - Added
bb_plot_year_recruitment()
. - Added
bb_plot_month_survival()
. - Added
bb_plot_year_survival()
. - Added
bb_plot_month()
.
- Added a
NEWS.md
file to track changes to the package.